End-side cloud collaborative task unloading method and device, equipment and medium

Abstract

The invention relates to the technical field of wireless communication, and discloses an end-side cloud collaborative task unloading method, device and equipment and a medium. The method comprises the following steps: respectively acquiring task sets of local processing of terminal equipment, unloading from the terminal equipment to an edge server for processing and unloading from the terminal equipment to a cloud server for processing, constructing a first task processing model for representing task processing of the terminal equipment, and constructing a second task processing model for representing task processing of the edge server, and constructing a third task processing model used for representing cloud server processing tasks, combining the first task processing model, the second task processing model and the third task processing model to form a target task processing model used for representing the task unloading problem in the end-side cloud cooperation scene, and taking the minimum execution energy consumption sum of all tasks as a target. And solving the target task processing model to obtain a task unloading strategy of the task unloading problem so as to improve the service quality of the task unloading problem in the end-side cloud collaborative scene.

Description

Technical Field

[0001] The present invention relates to the field of wireless communication technologies, and in particular to a method, device, equipment and medium for task offloading in edge-cloud collaboration. Background Art

[0002] After entering the era of Internet of Things (IoT), the number of mobile devices has grown exponentially, resulting in a large amount of data generation. At the same time, cloud computing has developed into a dominant computing paradigm that can provide on-demand and powerful computing capabilities for different applications. However, with the continuous increase in the amount of data, cloud computing faces major challenges in meeting the latency and privacy requirements in real-time applications, such as autonomous driving, real-time video analysis, and remote healthcare. Sending all data to the cloud for training or inference not only creates uncontrollable delays in data transmission and processing, but also increases the risk of privacy leakage and seriously affects the quality of service (Quality of Server, QoS).

[0003] At this time, mobile edge computing was proposed, that is, deploying computing nodes in areas close to smart terminals, with certain computing, storage and other functions. Since users are close to the computing nodes, data transmission delays can be effectively reduced, energy consumption can be reduced, and user experience can be effectively improved. Task processing efficiency can be improved by dynamically utilizing the collaborative computing of local devices, cloud servers and edge servers.

[0004] However, the heterogeneity of edge servers will bring a series of management problems: edge server computing nodes may fail during the execution of tasks due to hardware, software or other reasons. Usually, appropriate fault-tolerant technologies are used to improve the reliability of cloud data centers, so that these tasks can continue to execute from the failure point after the computing node failure is repaired, but the recovery process of computing node failure still takes a certain amount of time, which will inevitably affect the performance of the executed tasks. Summary of the invention

[0005] The purpose of the present invention is to provide a method, device, equipment and medium for task offloading in end-edge-cloud collaboration, which can improve the service quality of task processing problems in end-edge-cloud collaboration scenarios.

[0006] To solve the above technical problems, an embodiment of the present invention provides a task offloading method for device-edge-cloud collaboration, comprising the following steps:

[0007] Respectively obtain a first task set processed locally by the terminal device, a second task set offloaded from the terminal device to the edge server for processing, and a third task set offloaded from the terminal device to the cloud server for processing;

[0008] Constructing a first task processing model for characterizing the tasks processed by the terminal device according to the execution time and execution energy consumption of each task in the first task set;

[0009] Constructing a second task processing model for characterizing the edge server processing tasks according to the execution time and execution energy consumption of each task in the second task set, wherein the execution time of each task in the second task set includes the repair time of a fault occurring during the execution of the task;

[0010] Constructing a third task processing model for characterizing the tasks processed by the cloud server according to the execution time and execution energy consumption of each task in the third task set;

[0011] The execution time of each task in the first task set, the second task set, and the third task set is determined according to the execution resources pre-allocated thereto by the terminal device, the edge server, and the cloud server, respectively;

[0012] The first task processing model, the second task processing model and the third task processing model are combined to form a target task processing model for characterizing the task offloading problem in the end-edge-cloud collaboration scenario;

[0013] With the goal of minimizing the sum of execution energy consumption of all tasks, the target task processing model is solved and the task offloading strategy of the task offloading problem is obtained.

[0014] In some optional embodiments, the time of each failure of the edge server is independent of each other and obeys a Poisson process with a parameter of λ; the repair time of each failure is also independent of each other and obeys a Poisson process with a parameter of μ;

[0015] Among them, λ and μ are both greater than 0.

[0016] In some optional embodiments, the repair time of a fault occurring during the execution of each task in the second task set is expressed by the following formula:

[0017]

[0018] In the formula, represents the repair time of the failure that occurs during the execution of the task transmitted from terminal device i to the edge server, Indicates the execution time of the task.

[0019] In some optional embodiments, the execution time of each task in the second task set further includes a delay in transmitting the task from the corresponding terminal device to the edge server.

[0020] In some optional embodiments, the delay of transmitting the task from the corresponding terminal device to the edge server is expressed by the following formula:

[0021]

[0022] In the formula, represents the latency of the task transmitted from terminal device i to the edge server, D i represents the data volume of the task, B represents the channel bandwidth from the terminal device to the edge server, σ 2 represents the noise power, p i represents the transmission power, h i represents the channel gain, Indicates the interference caused by other terminal devices in the channel to terminal device i.

[0023] In some optional embodiments, the execution time of each task in the third task set includes a delay in transmitting the task from the corresponding terminal device to the edge server and a delay in transmitting the task from the edge server to the cloud server.

[0024] In some optional embodiments, the execution energy consumption of each task in the second task set includes the energy consumption of transmitting the task from the corresponding terminal device to the edge server and the energy consumption of the edge server executing the task;

[0025] The execution energy consumption of each task in the third task set includes the energy consumption of transmitting the task from the corresponding terminal device to the edge server and the energy consumption of transmitting the task from the edge server to the cloud server.

[0026] An embodiment of the present invention further provides a device-edge-cloud collaborative task offloading device, comprising:

[0027] A task acquisition module, used to respectively acquire a first task set processed locally by a terminal device, a second task set offloaded from the terminal device to an edge server for processing, and a third task set offloaded from the terminal device to a cloud server for processing;

[0028] A model building module, used to build a first task processing model for characterizing the terminal device processing tasks according to the execution time and execution energy consumption of each task in the first task set;

[0029] Constructing a second task processing model for characterizing the edge server processing tasks according to the execution time and execution energy consumption of each task in the second task set, wherein the execution time of each task in the second task set includes the repair time of a fault occurring during the execution of the task;

[0030] Constructing a third task processing model for characterizing the tasks processed by the cloud server according to the execution time and execution energy consumption of each task in the third task set;

[0031] The execution time of each task in the first task set, the second task set, and the third task set is determined according to the execution resources pre-allocated thereto by the terminal device, the edge server, and the cloud server, respectively;

[0032] The first task processing model, the second task processing model and the third task processing model are combined to form a target task processing model for characterizing the task offloading problem in the end-edge-cloud collaboration scenario;

[0033] The task decision module is used to solve the target task processing model with the goal of minimizing the sum of the execution energy consumption of all tasks, and obtain the task offloading strategy for the task offloading problem.

[0034] An embodiment of the present invention also provides a computer device, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the above-mentioned end-edge-cloud collaborative task offloading method.

[0035] An embodiment of the present invention further provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the above-mentioned task offloading method for end-edge-cloud collaboration.

[0036] The task offloading method of device-edge-cloud collaboration provided by the present invention has at least the following beneficial effects:

[0037] In the end-edge-cloud collaborative computing system, the system will decide whether the task is executed locally or by offloading to the edge server or cloud server through offloading decisions. Then, models are established for the local execution of the task, the edge server execution, and the cloud server execution, thereby forming a model of the task offloading problem in the end-edge-cloud collaborative scenario. This model takes into account the latency and energy consumption of tasks locally, on edge servers, and on cloud servers, and is more in line with the actual task offloading situation. The task processing latency includes the repair time of faults that occur during the execution of the edge server when executing the task. Based on this, with the goal of minimizing the sum of the execution energy consumption of all tasks, the model of the task offloading problem in the end-edge-cloud collaborative scenario is solved, and the system task execution energy consumption is minimized while considering the impact of factors such as edge server failures and fault recovery on the task processing latency.

[0038] At the same time, by allocating edge server resources in real time, it is ensured that after a failure occurs, the task is completed within the maximum allowed execution period with a certain execution success rate, reducing the problem of task execution failure and ensuring service quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] One or more embodiments are exemplarily described by the pictures in the corresponding drawings, and these exemplary descriptions do not constitute limitations on the embodiments.

[0040] Figure 1 is a flowchart of a task offloading method for end-edge-cloud collaboration provided according to an embodiment of the present invention;

[0041] Figure 2 is a schematic diagram of a device-edge-cloud collaboration scenario provided according to an embodiment of the present invention;

[0042] Figure 3 is a schematic diagram of a task offloading device for end-edge-cloud collaboration provided according to an embodiment of the present invention;

[0043] Figure 4 It is a structural schematic diagram of a computer device provided according to an embodiment of the present invention. DETAILED DESCRIPTION

[0044] In order to make the purpose, technical scheme and advantages of the embodiments of the present invention clearer, the embodiments of the present invention will be described in detail below in conjunction with the accompanying drawings. However, it will be appreciated by those skilled in the art that in the embodiments of the present invention, many technical details are proposed in order to enable the reader to better understand the present invention. However, even without these technical details and various changes and modifications based on the following embodiments, the technical scheme claimed in the present invention can be implemented. The division of the following embodiments is for the convenience of description and should not constitute any limitation on the specific implementation of the present invention. The various embodiments can be combined and referenced with each other without contradiction.

[0045] An embodiment of the present invention relates to a task offloading method for end-edge-cloud collaboration. The implementation details of the task offloading method for end-edge-cloud collaboration of this embodiment are described in detail below. The following content is only the implementation details provided for easy understanding and is not necessary for implementing this solution.

[0046] The specific process of the task offloading method of the device-edge-cloud collaboration in this embodiment can be as follows: Figure 1 As shown, including:

[0047] Step 101 , respectively obtain a first task set to be processed locally on a terminal device, a second task set to be offloaded from the terminal device to an edge server for processing, and a third task set to be offloaded from the terminal device to a cloud server for processing.

[0048] Specifically, since cloud servers are rich in resources, they can handle tasks with large amounts of data, but cloud servers have difficulty handling some latency-sensitive tasks. Edge servers are deployed at base stations, close to user devices, and can solve latency-sensitive tasks, but resources are relatively scarce. Therefore, considering an edge-cloud collaborative computing system, the system will decide whether the task is executed locally or offloaded to the edge server or cloud server through the base station through offloading decisions, such as Figure 2 The scenario of end-edge-cloud collaborative computing.

[0049] Therefore, in this scenario, a first task set processed locally on the terminal device, a second task set offloaded from the terminal device to the edge server for processing, and a third task set offloaded from the terminal device to the cloud server for processing can be obtained respectively.

[0050] Step 102: construct a target task processing model for characterizing task processing offloading in the end-edge-cloud collaboration scenario.

[0051] Specifically, assuming that K terminal devices N = {N1, N2, ..., N i ,...,N k}, in the entire time T, at intervals ΔT, a user publishes a task The set of tasks in time slot ΔT can be In other words, there are TK / ΔK tasks in the whole time. Each task can use the triple {D i ,C i ,T max} represents its characteristics, where D i Indicates the amount of data for the task (in bits), C i Indicates the number of CPU cycles required for each bit of the computing task, T max Represents the maximum execution time of the task, with a l ,a m ,a c Respectively represent execution on the terminal device, edge server, and cloud server, then there is a l ,a m ,a c ∈{0,1} and a l +a m +a c =1.

[0052] First, according to the execution time and execution energy consumption of each task in the first task set, a first task processing model for characterizing the terminal device processing tasks is constructed, as described in detail as follows:

[0053] The first task set executed locally by the terminal device is Terminal device Ni The local CPU computing resources are f i,L (in cycles / second), the execution time of each task in the first task set is expressed as:

[0054]

[0055] The execution energy consumption of each task in the first task set is expressed as:

[0056]

[0057] Where κ depends on the energy coefficient of the chip structure.

[0058] Then, the total energy consumption of the first task set executed locally by the terminal device is:

[0059]

[0060] Then, according to the execution time and execution energy consumption of each task in the second task set, a second task processing model for characterizing the edge server processing tasks is constructed, as described in detail as follows:

[0061] The second task set offloaded to the edge server is Terminal device N i The channel bandwidth to the edge server is B, σ 2 represents the noise power, p i represents the transmission power, h i represents the channel gain. Since the edge server is close to the base station, the time for data transmission to the edge server is negligible. Then the uplink data transmission rate R offloaded to the edge server is i It can be expressed as:

[0062]

[0063] In the formula, represents the interference caused by other terminal devices in the channel to terminal device i, k off\{i} represents all other terminal devices in the channel except terminal device i, p k represents the transmission power of the kth terminal device among all other terminal devices, h k Represents the channel gain of the kth terminal device among all other terminal devices.

[0064] Then the delay of task transmission from the corresponding terminal device to the edge server is:

[0065]

[0066] Assume that the total resources of the edge server are f M , assigned to the task posted by user i The CPU resource is f i,M , then the execution time of the edge server for the task is:

[0067]

[0068] In order to model the task offloading problem considering edge server failure in the edge-cloud collaborative environment, the following assumptions are made:

[0069] 1) Assume that the failures of edge servers are independent of each other and follow a Poisson process with parameter λ>0, that is, the interval time series of edge server failures are independent of each other and follow a negative exponential distribution with parameter λ.

[0070] 2) Assume that all faults are repairable, and the repair time of each fault is independent of each other and follows a negative exponential distribution with parameter μ. That is, the fault repair rate is μ.

[0071] use Indicated in The total number of failures of an edge server during the execution of the task, is a random variable, and The probability is:

[0072]

[0073] The expected number of failures during task execution is:

[0074]

[0075] Using RT (k) It represents the repair time of the kth (k=1,2,…) failure of an edge server, and its expected value is E(RT (k) )=1 / μ, where μ is the fault repair rate of the faulty computing node, using Indicates that the faulty edge server is The total expected repair time for a failure in time is:

[0076]

[0077] Since the amount of downlink data is small and the transmission time is negligible, the task The total delay of failure during execution is and The task execution offloaded to the edge server is The total time is:

[0078]

[0079] The energy consumption of transmitting the task from the corresponding terminal device to the edge server is:

[0080]

[0081] Since edge servers are not as stable as cloud servers, the energy consumption of each CPU cycle of the server is ρ c , then the energy consumption of the edge server to perform tasks is:

[0082]

[0083] Task The total energy required is:

[0084]

[0085] The total energy consumption of the second task set on the edge server is:

[0086]

[0087] Next, according to the execution time and execution energy consumption of each task in the third task set, a third task processing model for characterizing the cloud server processing tasks is constructed, as described in detail as follows:

[0088] The task set offloaded to the cloud server is in, The latency of task transmission from the corresponding terminal device to the edge server is:

[0089]

[0090] The latency from edge server to cloud server is:

[0091]

[0092] Where W i Indicates the communication capacity from edge server to server.

[0093] Task The latency of offloading to the cloud server is:

[0094]

[0095] The cloud server allocates tasks based on user i’s pre-subscription The CPU resource is f i,C , then the execution time of the cloud server for the task is:

[0096]

[0097] Since the amount of downlink data is small and the transmission time is negligible, the task The total execution delay is and The sum of the tasks is offloaded to the cloud server for execution The total time is:

[0098]

[0099] The energy consumption of transmitting tasks from the corresponding terminal devices to the cloud server is:

[0100]

[0101] Since the cloud server has a stable power supply, the execution energy consumption is not considered. The total energy required is:

[0102]

[0103] The total energy consumption of the task set on the cloud server is:

[0104]

[0105] Finally, the first task processing model, the second task processing model and the third task processing model are combined to form a target task processing model for characterizing task processing offloading in the end-edge-cloud collaboration scenario.

[0106] Step 103, taking the minimum sum of execution energy consumption of all tasks as the goal, solving the target task processing model, and obtaining the task offloading strategy for the task offloading problem.

[0107] The optimization objectives and constraints are as follows:

[0108]

[0109] Based on this, the task offloading strategy of the task offloading problem is solved.

[0110] In this embodiment, in the end-edge-cloud collaborative computing system, the system will decide whether the task is executed locally or by offloading to the edge server or cloud server through offloading decisions. Then, models are established for the local execution of the task, the edge server execution, and the cloud server execution, respectively, so as to form a model of the task offloading problem in the end-edge-cloud collaborative scenario. This model takes into account the delay and energy consumption of the task locally, on the edge server, and on the cloud server, and is more in line with the actual task offloading situation. The task processing delay includes the repair time of the fault that occurs during the execution of the edge server when the task is executed. Based on this, the model of the task offloading problem in the end-edge-cloud collaborative scenario is solved with the goal of minimizing the sum of the execution energy consumption of all tasks. Under the premise of considering the impact of factors such as edge server failure and fault recovery on the task processing delay, the system task execution energy consumption is minimized. At the same time, by allocating edge server resources in real time, it is ensured that the task is completed before the maximum allowed execution period with a certain execution success rate after the failure occurs, reducing the problem of task execution failure and ensuring the service quality.

[0111] The step division of the various methods above is only for clear description. When implemented, they can be combined into one step or some steps can be split and decomposed into multiple steps. As long as they include the same logical relationship, they are all within the protection scope of the present invention. Adding insignificant modifications or introducing insignificant designs to the algorithm or process without changing the core design of the algorithm and process are all within the protection scope of the invention.

[0112] Another embodiment of the present invention relates to a task offloading device for end-edge-cloud collaboration. The implementation details of the task offloading device for end-edge-cloud collaboration of this embodiment are described in detail below. The following content is only for the convenience of understanding the implementation details provided, and is not necessary for the implementation of this solution. The schematic diagram of the task offloading device for end-edge-cloud collaboration of this embodiment can be as follows: Figure 3 As shown, it includes: a task acquisition module 301, a model building module 302 and a task decision module 303.

[0113] Specifically, the task acquisition module 301 is used to respectively acquire a first task set processed locally by the terminal device, a second task set offloaded from the terminal device to the edge server for processing, and a third task set offloaded from the terminal device to the cloud server for processing.

[0114] A model building module 302, configured to build a first task processing model for characterizing the terminal device processing tasks according to the execution time and execution energy consumption of each task in the first task set;

[0115] Constructing a second task processing model for characterizing the edge server processing tasks according to the execution time and execution energy consumption of each task in the second task set, wherein the execution time of each task in the second task set includes the repair time of a fault occurring during the execution of the task;

[0116] Constructing a third task processing model for characterizing the tasks processed by the cloud server according to the execution time and execution energy consumption of each task in the third task set;

[0117] The execution time of each task in the first task set, the second task set, and the third task set is determined according to the execution resources pre-allocated thereto by the terminal device, the edge server, and the cloud server, respectively;

[0118] The first task processing model, the second task processing model and the third task processing model are combined to form a target task processing model for characterizing task processing problems in the end-edge-cloud collaboration scenario.

[0119] The task decision module 303 is used to solve the target task processing model with the goal of minimizing the sum of the execution energy consumption of all tasks, and obtain the task offloading strategy for the task offloading problem.

[0120] It is not difficult to find that this embodiment is a device embodiment corresponding to the above method embodiment, and this embodiment can be implemented in conjunction with the above method embodiment. The relevant technical details and technical effects mentioned in the above embodiment are still valid in this embodiment, and in order to reduce repetition, they are not repeated here. Accordingly, the relevant technical details mentioned in this embodiment can also be applied in the above embodiment.

[0121] It is worth mentioning that all modules involved in this embodiment are logic modules. In practical applications, a logic unit can be a physical unit, a part of a physical unit, or a combination of multiple physical units. In addition, in order to highlight the innovative part of the present invention, this embodiment does not introduce units that are not closely related to solving the technical problem proposed by the present invention, but this does not mean that there are no other units in this embodiment.

[0122] Another embodiment of the present invention relates to a computer device, such as Figure 4 As shown, it includes: at least one processor 401; and a memory 402 that is communicatively connected to the at least one processor 401; wherein the memory 402 stores instructions that can be executed by the at least one processor 401, and the instructions are executed by the at least one processor 401 so that the at least one processor 401 can execute the end-edge-cloud collaborative task offloading method in the above-mentioned embodiments.

[0123] Among them, the memory and the processor are connected in a bus manner, and the bus may include any number of interconnected buses and bridges, and the bus connects various circuits of one or more processors and memories together. The bus can also connect various other circuits such as peripherals, voltage regulators, and power management circuits, which are well known in the art and are therefore not further described herein. The bus interface provides an interface between the bus and the transceiver. The transceiver can be one element or multiple elements, such as multiple receivers and transmitters, providing a unit for communicating with various other devices on a transmission medium. The data processed by the processor is transmitted on a wireless medium via an antenna, and further, the antenna also receives data and transmits the data to the processor.

[0124] The processor is responsible for managing the bus and general processing, and can also provide various functions, including timing, peripheral interfaces, voltage regulation, power management, and other control functions. Memory can be used to store data used by the processor when performing operations.

[0125] Another embodiment of the present invention relates to a computer-readable storage medium storing a computer program, which implements the above method embodiment when executed by a processor.

[0126] That is, those skilled in the art can understand that all or part of the steps in the above-mentioned embodiment method can be completed by instructing the relevant hardware through a program, and the program is stored in a storage medium, including a number of instructions for a device (which can be a single-chip microcomputer, chip, etc.) or a processor (processor) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (Read-Only Memory, referred to as: ROM), random access memory (Random Access Memory, referred to as: RAM), disk or optical disk and other media that can store program codes.

[0127] Those skilled in the art will appreciate that the above embodiments are specific embodiments for implementing the present invention, and in actual applications, various changes may be made thereto in form and detail without departing from the spirit and scope of the present invention.

Claims

1. A task offloading method for device-edge-cloud collaboration, characterized in that: The method comprises: Respectively obtain a first task set processed locally by the terminal device, a second task set offloaded from the terminal device to the edge server for processing, and a third task set offloaded from the terminal device to the cloud server for processing; Constructing a first task processing model for characterizing the tasks processed by the terminal device according to the execution time and execution energy consumption of each task in the first task set; Constructing a second task processing model for characterizing the edge server processing tasks according to the execution time and execution energy consumption of each task in the second task set, wherein the execution time of each task in the second task set includes the repair time of a fault occurring during the execution of the task; Constructing a third task processing model for characterizing the tasks processed by the cloud server according to the execution time and execution energy consumption of each task in the third task set; The execution time of each task in the first task set, the second task set, and the third task set is determined according to the execution resources pre-allocated thereto by the terminal device, the edge server, and the cloud server, respectively; Combining the first task processing model, the second task processing model, and the third task processing model to form a target task processing model for characterizing task processing offloading in the end-edge-cloud collaboration scenario; With the goal of minimizing the sum of execution energy consumption of all tasks, the target task processing model is solved and the task offloading strategy of the task offloading problem is obtained.

2. The task offloading method for edge-cloud collaboration according to claim 1 is characterized in that: The time when each failure of the edge server occurs is independent of each other and obeys a Poisson process with a parameter of λ; the time when each failure is repaired is also independent of each other and obeys a Poisson process with a parameter of μ; Among them, λ and μ are both greater than 0.

3. The task offloading method for edge-cloud collaboration according to claim 2 is characterized in that: The repair time of a fault occurring during the execution of each task in the second task set is expressed by the following formula: In the formula, represents the repair time of a failure that occurs during the execution of a task transmitted from terminal device i to the edge server, Indicates the execution time of the task.

4. The task offloading method for edge-cloud collaboration according to claim 1 is characterized in that: The execution time of each task in the second task set also includes the delay of transmitting the task from the corresponding terminal device to the edge server.

5. The task offloading method for edge-cloud collaboration according to claim 4 is characterized in that: The delay of transmitting the task from the corresponding terminal device to the edge server is expressed by the following formula: In the formula, represents the latency of the task transmitted from terminal device i to the edge server, D i represents the data volume of the task, B represents the channel bandwidth from the terminal device to the edge server, σ 2 represents the noise power, p i represents the transmission power, h i represents the channel gain, Indicates the interference caused by other terminal devices in the channel to terminal device i.

6. The task offloading method for edge-cloud collaboration according to claim 1 is characterized in that: The execution time of each task in the third task set includes the delay of transmitting the task from the corresponding terminal device to the edge server and the delay of transmitting the task from the edge server to the cloud server.

7. The task offloading method for edge-cloud collaboration according to any one of claims 1 to 6, characterized in that: The execution energy consumption of each task in the second task set includes the energy consumption of transmitting the task from the corresponding terminal device to the edge server and the energy consumption of the edge server executing the task; The energy consumption of each task in the third task set includes the energy consumption of transmitting the task from the corresponding terminal device to the edge server and the energy consumption of transmitting the task from the edge server to the cloud server.

8. A task offloading device for end-edge-cloud collaboration, characterized in that: The device comprises: A task acquisition module, used to respectively acquire a first task set processed locally by a terminal device, a second task set offloaded from the terminal device to an edge server for processing, and a third task set offloaded from the terminal device to a cloud server for processing; A model building module, used to build a first task processing model for characterizing the terminal device processing tasks according to the execution time and execution energy consumption of each task in the first task set; Constructing a second task processing model for characterizing the edge server processing tasks according to the execution time and execution energy consumption of each task in the second task set, wherein the execution time of each task in the second task set includes the repair time of a fault occurring during the execution of the task; Constructing a third task processing model for characterizing the tasks processed by the cloud server according to the execution time and execution energy consumption of each task in the third task set; The execution time of each task in the first task set, the second task set, and the third task set is determined according to the execution resources pre-allocated thereto by the terminal device, the edge server, and the cloud server, respectively; Combining the first task processing model, the second task processing model, and the third task processing model to form a target task processing model for characterizing task processing problems in a device-edge-cloud collaboration scenario; The task decision module is used to solve the target task processing model with the goal of minimizing the sum of the execution energy consumption of all tasks, and obtain the task offloading strategy for the task offloading problem.

9. A computer device, characterized in that: include: at least one processor; And, a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the end-edge-cloud collaborative task offloading method as described in any one of claims 1 to 6.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the task offloading method for end-edge-cloud collaboration according to any one of claims 1 to 6 is implemented.

Images